Next Gen PSF - Blog/News

The Future of AI in Accounting: In Conversation with Hugh Shields (ICAS)

As part of the Next Generation Professional Service Firms Project, Hilary Smyth-Allen talks to Hugh Shields, Head of AI and Technology at ICAS, about his thoughts on the future of AI in the accountancy sector and the questions that this will raise for professional services firms.

It’s early on a Friday morning, central London and I can already tell that this is going to be one of those fascinating wide-ranging conversations and that the normal allotted hour for the formal research interview probably isn’t going to cut it. A rapid exchange of text and fleeting conversations reveal that I am in the right place, but Hugh’s a bit behind due to an over-run of his morning training session in preparation for a European athletics meet next month where he is representing GB. But in the meantime, I could read Hugh’s vision paper on the accounting sector in 2050, which arrives soon after via email. Our conversation begins….

Hilary: You’re fairly new in position with ICAS, I believe. Can you tell me a bit more about you and the role?

Hugh: That’s right. In early March, ICAS launched a new agenda called Technology, Trust and Talent. I’m the lead for the technology strand within that, being announced as the new Head of AI and Technology for the institution at that launch event. This is a part-time role that I am doing alongside my primary employment as a Global Director at Huawei Technologies.

Hugh: I have a deep interest in AI generally, the technological capabilities and so forth. But fundamentally it’s about business processes. AI isn’t goal in itself unless anchored in genuine understanding of the business processes.

Let me give you a simple example from the housing association sector, where I volunteer as a Non-executive Director and a simple AI application in form of a chatbot to help address tenants’ issues, such as ”I’ve got a plumbing problem”. So they think they’re going to have a chat bot to help with the efficiency. Well the business process part of it might say have you analysed why you’re getting all these plumbing problems from this area because you get far more from this area than you do from anywhere else round the country. There’s something systematic going wrong, which — by the way — the data could tell you; structured data would be able to tell you that, so there’s a sort of a technology angle there, or there is a technology angle there. But then you could go underneath and say, oh actually there’s this dodgy plumbing firm and they’re behind all these problems, so we could go and fix all these houses now, and then you won’t get the call coming through to the chat bot at all.

Hilary: So you don’t need a chat bot then?

Hugh: That’s right; you have to understand business process, this is, because to me, I mean I'm not a technologist, but I basically now assume that anything is possible from a machine, if not necessarily today, then quite soon. My experience of working on AI project with the top AI scientists, I mean some of the very best people in the world is that every question I ask about the art of the possible is answered with “yes, we can do that”… basically anything I ask is possible in terms of technological AI capability.

So my role and benefit to these very bright technologists is to ask the right business processing questions to enable relevant and appropriate deployment of technologies. But the constraint that nearly always exists in the middle is access to the right data.

Hilary: Can you expand a bit more on what the remit of your role is with ICAS please?

Hugh: Sure. The role that I have is to help member firms adopt new technologies, perhaps AI but perhaps other things. It’s to help inform policy debate about AI, which is why your research at Sheffield University is an interesting project to engage with for this work. I have a particular angle that I'm coming from with all of this, which is the technological singularity, is basically the idea that at some point, maybe 20 years away, maybe 30 years away, but pretty close, robots can basically do every job that we can see today. I mean everything. Now, if you have that as a starting point then you start to think, gosh, what does this mean? For me as an individual, for my company I work for, for society as a whole, for mankind? Is this debate happening? I don’t think that I, at least not on all the different levels that it ought to be? Technology is moving so fast but we mustn’t miss the debate.

There are these are big questions that need addressing because in 30 years time, let’s say, you basically have a situation where robots are doing all the work, and the humans can be sitting on the beach drinking their gin and cocktail, or not, and this is the question. What do you want to do? How do you want to play it? Do you need companies? I mean if robots are doing all the work, all of it, do you need a board of directors skimming off all the profits? At some point the robots could literally be doing everything. I believe that we are capable of doing away with human accountants as we know it.

Hilary: But what about judgement? Isn’t that the bastion of the professions and humans?

Hugh: There is nothing technologically speaking, I don’t think, to stop machines doing all accounting. Right now. That’s how I see it. People say, well what about judgements? To which I respond to ask what about them? Take an example, say the complexities of what to provide for in a bank’s accounts when it receives a piece of litigation. How much should they provide for that? Yes, it’s complicated, but you could load a machine up with all the cases, all the weightings, far more evidence, far quicker and give you a much better answer than any human.

Hilary: But the weak link being the data bit again?

Hugh: Well, yes. But the point is that there’s nothing stopping us loading the machine up with the data. That can be done. What is human judgement then? Well if you ask an audit partner, what do you think this should be, they’ll likely talk through their previous experiences of it. But isn’t that the same as giving the past in the form of data to a machine and getting a better, faster, more evidenced answer to the question about the provision?

I took this role with ICAS to have exactly these debates and support the profession in adjusting to the new ways in which work can be done. Sure we don’t have the data in the right ways as yet, but that’s a temporary blip in the direction of travel which is one where technology challenges the business model entirely. I hope to be able to assist in leading our profession through this transformation as the first chartered accountancy professional body.

This blog was created, with permission, from stakeholder engagement activities conducted as part of the Industrial Strategy Challenge Fund Next Generation Services project, awarded to the University of Sheffield, University of the Arts London, Lancaster University, and the University of Manchester. The research programme seeks to understand and address the challenges facing mid-sized professional services firms in law and accountancy in adopting and implementing AI technologies and is supported by the ESRC. Over the coming months the Next Generation PSF project team will be working with a number of legal and accounting firms to explore how they can approach the AI challenge. If you are interested in finding out more, or want to have your firm participate in one of the sessions facilitated by the Next Generation PSF project team, contact us at

Posted 19th July 2019

AI for CFOs

There is little question about the importance of Chief Finance Officers, where the responsibility for managing the financial actions of a business is as much about tomorrow as it is about today. When it comes to thinking about the role of AI in professional services firms it is therefore critical that the CFO is onboard. However, for many CFOs the reality is that there is a lack of understanding about the applications and potential of AI (and arguably technology more generally). As such, the NextGenPSF team have been contributing to a series of events with the CFO forum to raise awareness around the potential of AI and the need to embrace AI investment.

Recent research by Grant Thornton and CFO research talks about how the digital transformation journey will also require CFOs to alter their mindset when it comes to technology investments, and this represents a particular challenge for professional services firms. With the majority of firms still functioning as partnerships, there is often little incentive for partners to invest in technology when it directly impacts on their remuneration and bonus. That said there is growing recognition about the importance of not just investing in technology but getting it right.

There is a need to move beyond the often negative attitudes towards investing in AI, which is perpetuated because AI for many is still a largely unknown quantity. If, however, the question of investing in AI is addressed as a strategic concern the tone of the conversation can be fundamentally different. The decision to invest AI is more likely to overcome the inevitable return on investment question, if it either addresses a problem in the business or enables the business to realise a growth opportunity. As such the decision to invest in, and adopt, AI needs to be understood as a business model question.

Over the coming months the Next Generation PSF project team will be working with a number of legal and accounting firms to explore how they can approach the AI challenge. If you are interested in finding out more, or want to have your firm participate in one of the sessions facilitated by the Next Generation PSF project team, contact us at

Posted 24th June 2019

Next Gen Services Project Assembly

On 14th May our NextGenPSF project team participated in an event for all funding recipients under the Next Generation Services (NGS) Industrial Strategy Challenge Fund (ICSF) priority. The gathering brought together representatives from across the academically focued research projects as well as the more appliedcollaborative research and development (CRD) projects.

In framing the day Matt Hotson (CFO UK & International, RSA Insurance Group) reflected on the transformative potential of AI to the legal, accounting and insurance sectors currently prioritised under the ISCF. As well as highlighting the technical possibilities of new technologies associated under the grand challenge of AI and the data economy, CFO flagged how questions of people and strategy as critical to future competitiveness.

Tim Vorley provided an overview of the NextGenPSF project (see the above video), alongside John Armour (University of Oxford) and Alistair Milne (Loughborough University) . The presentations highligted clear links across the projects relating to the blurring of boundaries at the frontiers of the professions, as well as emphasising the importance of business models and business model innovation. Critically for the NextGenPSF project we regognised that the threat to mid-tier legal and accounting firms comes from above (i.e. the inverstment of the Big 4 and Magic Circle) as well as below (i.e. the start-ups disrupting through accounting, audit and legal-tech).

Another important outcome from the day was the emphasis on data. While frequently referred to as the new oil, it is imperative that the critical and value added of the refining process not be overlooked. Jeni Tennison (CEO, Open Data Institute) raised the importance of data as infrastructure and the need to ensure that the appropriate institutions are in place to ensure the future competitiveness of data intensive business and sectors. This challenge will be addressed directly by the NextGenPSF project.

Posted 14th May 2019

Cultural Challenges: AI in Professional services Firms

A major challenge facing the implementation of artificial intelligence into legal professional services is an inertia to embrace change. Many legal services have been dominated by business models dating back centuries where modes of practice are established and rooted in convention and norms. This has in part led to a reticence in some areas to adopt AI to complement and extend current practices.

The opportunities afforded by the careful integration of AI within legal professional services firms are manifold. For example, the automation of labour-intensive tasks that can free up individuals to concentrate on more demanding issues, assisting due diligence, and the ability to process and analyse large amounts of data. There is a plethora of technologies which have filtered into the legal profession including Kira, LitIQ and Neotalogic amongst others to help with documentation processing, legal analytics, and even to assist prediction.

The successful integration of AI carries benefits to both the competitiveness and productivity of firms as well as to the wider UK economy. The role of business model innovation in this regard is paramount to creating new efficiencies, and reimagining traditional practices to remain competitive and sustainable in the professional services landscape. Business model innovation shifts the focus from technology as the sole driver of innovation to a more nuanced approach. This approach focuses on the disruption of the configuration of business activities, and seeks to innovate throughout all aspects of how a firm is organised, engages with its clients, and shares its services.

Creating an economy that harnesses these AI technologies, coupled with big data and the digital economy, are seen as one of the great opportunities of our age, especially for early movers and innovators within key industries. Whilst issues of regulation and the future of work remain important policy considerations, it is incumbent on firms and stakeholders to ensure the future of the UK professional services sector. Through the anticipation of future scenarios and considering the business model more creatively as a dynamic facet of legal services practice, the future of AI may find a more certain home.

Posted 4th May 2019

Employment Outlook on the Future of Work

On 25th April at the Work Foundation in London, James Faulconbridge used insights from the Next Generation PSF project to contribute to a panel discussion on the OECD’s 2019 Employment Outlook report which focuses on the future of work. The report highlights the need to focus more on working with new technologies such as AI and less on the jobs that might be lost –the OECD’s research suggests it is how tasks associated with particular jobs will change that is of greatest significance. This aligns well with the initial findings of the Next Generation PSF project.

In his contribution, James focussed on the change process associated with AI adoption. He highlighted the importance of examining what it means to work with AI, and how this applies to professionals as much as lower skilled workers. James highlighted the need to examine the way firms handle the transitions involved, in our project this relating to the way the work practices of accountants and lawyers change. In particular this raises questions about how firms can enable the transitions involved through training, support for learning in practice as part of day-to-day work and active management of the structures of teams and work relationships.

In the discussions an important debate developed about the need to focus on what working with AI will mean for careers and different generations of workers. In line with the focus of the Next Generation PSF project, it was highlighted that a ‘whole system’ approach is needed that considers the skills earlier careers workers have and need to develop through training, how more experienced workers will transition, and how new roles, tasks and contributions will emerge within occupations as AI fulfils certain roles and fees-up workers to add-value in other ways. It was agreed that trying to predict such changes is futile, and instead we need toolkits and approaches that help the process of change to be understood and managed effectively, by firms, workers, regulators and other stakeholders.

The Next Generation PSF project will be focussing on all of these issues in the forthcoming 24 months as we seek to develop a better understanding of how accounting and law firms can respond to the transitions associated with the growing role and potential of AI.

OECD Employment Outlook report available HERE

Posted 26th April 2019

Launching Next Generation Services

Following the announcement on 28 November 2018 from BEIS about funding for tech-driven legal and accounting boost productivity, we are delighted that the University of Sheffield led project was funded . Our project is primarily concerned with the human (i.e. non-technical) aspects of implementing new AI technologies in mid-size accounting and legal services firms. As well as analysing the potential barriers to adopting AI-based, the project is primarily concerned with harnessing the disruptive potential of AI and stimulating business model innovation for growth. With AI set to transform professional services world-leading legal services , our project is about supporting mid-market firms to design approached to strengthen their position in the market. The slides below provide an overview of the NextGenPSF project.

UKRI meeting.pdf

The other 2 projects funded were:

Professor Milne’s research will explore the implications for the insurance industry of this wave of new digital technologies, with the support of many of the UK’s leading insurance companies. They will identify and map the range of opportunities for AI based innovation in business processes and business models, across underwriting and risk analytics, claims processing and customer engagement. They will examine, through engagement with industry on business opportunities and challenges and through a range of case studies, the barriers to adoption and the enablers of change.

Professor Armour’s research will look into the use of AI in the legal system. His research seeks to identify how constraints on the implementation of AI in legal services can be relaxed to unlock its potential for good. As well as governing economic order, the legal system is more fundamentally a structure for social order. As a result the stakes for AI’s implementation in UK legal services are high. If mishandled, it could threaten both economic success and governance more generally.

Posted 11th January 2019